Temporarily crash the GPU process when unexpected failures happen.

Temporarily crash the GPU process when unexpected failures happen.

There is an ongoing issue with failures on Mac OSX causing crashes
on the renderer side. All we know so far is that it has something to do
with failures with our usage of GL_COMMANDS_COMPLETED_CHROMIUM. We
suspect it has something to do with multiple GPUs but it is hard to
tell for sure. This will at least let us know what is going on for
the GPU side.

BUG=467678
Committed: https://crrev.com/27cfeeeaa3faee230d2d526b13a3b0dee056022e
Cr-Commit-Position: refs/heads/master@{#339703}

[David Yen, 2015-07-20 22:53:48]

dyen@chromium.org changed reviewers:
+ zmo@chromium.org

[David Yen, 2015-07-20 23:00:32]

I am adding these crashes so we can know what is causing the GPU to fail for
these GL_COMMANDS_COMPLETED_CHROMIUM failures. We don't get the crash line
though do we? Is there any way to embed extra data into the crash info?

[Zhenyao Mo, 2015-07-20 23:01:56]

On 2015/07/20 22:53:49, David Yen wrote:

LGTM

[Zhenyao Mo, 2015-07-20 23:21:47]

On 2015/07/20 23:01:56, Zhenyao Mo wrote:
> On 2015/07/20 22:53:49, David Yen wrote:
> 
> LGTM

LGTM

[David Yen, 2015-07-20 23:24:21]

The CQ bit was checked by dyen@chromium.org

[commit-bot: I haz the power, 2015-07-20 23:24:39]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1242913003/60001

[commit-bot: I haz the power, 2015-07-21 00:50:09]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2015-07-21 00:50:10]

Try jobs failed on following builders:
  mac_chromium_rel_ng on tryserver.chromium.mac (JOB_FAILED,
http://build.chromium.org/p/tryserver.chromium.mac/builders/mac_chromium_rel_...)

[David Yen, 2015-07-21 16:32:56]

The CQ bit was checked by dyen@chromium.org

[David Yen, 2015-07-21 16:32:56]

The patchset sent to the CQ was uploaded after l-g-t-m from zmo@chromium.org
Link to the patchset: https://codereview.chromium.org/1242913003/#ps80001
(title: "Removed unit test temporarily for mac")

[commit-bot: I haz the power, 2015-07-21 16:33:41]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1242913003/80001

[commit-bot: I haz the power, 2015-07-21 18:28:17]

Committed patchset #5 (id:80001)

[commit-bot: I haz the power, 2015-07-21 18:29:21]

Patchset 5 (id:??) landed as
https://crrev.com/27cfeeeaa3faee230d2d526b13a3b0dee056022e
Cr-Commit-Position: refs/heads/master@{#339703}

[piman, 2015-07-21 19:17:05]

piman@chromium.org changed reviewers:
+ piman@chromium.org

[piman, 2015-07-21 19:17:06]

https://codereview.chromium.org/1242913003/diff/80001/gpu/command_buffer/serv...
File gpu/command_buffer/service/gles2_cmd_decoder.cc (right):

https://codereview.chromium.org/1242913003/diff/80001/gpu/command_buffer/serv...
gpu/command_buffer/service/gles2_cmd_decoder.cc:11626: CHECK(false) << "Unknown
error.";
drive-by FYI: you can use base::debug::Alias to make sure a particular stack
variable is not optimized, so you can observe it from the crash dump.